Exact calculation of the aggregate claims distribution in the individual life model by use of an n-layer model

Abstract

A recursive procedure is suggested for calculating the aggregate claims distribution (stop-loss premium) in the individual life model. The method which is based on the well-known de Pril algorithm results in both a considerably reduction of the number of arithmetic operations to be carried out and the number of data to be kept at each step of iteration. The problem of underflow/ overflow which may arise in case of a large number of policies is avoided by iterating in different layers and by suitably defining the transitions between adjacent layers. Thus the algorithm can be applied to a portfolio with an arbitrary number of policies.